Soil sampling planning in traceability studies by means of Experimental Design approaches

• The selection of sub-set of soil samples for setting up traceability models was performed.
• Geological features and several production variables were taken into account.
• Principal component analysis was used for the evaluation of soil characterization data.
• Experimental Design technique was used for soil selection.

The present research is part of a project dealing with the development of analytical methodologies mainly based on primary indicators, such as isotopic ratios of stable and radiogenic elements, for the authenticity and geographical traceability of oenological food, typical of the Modena district. In particular, considering the objective of establishing a food-territory link by means of these analytical indicators, it is straightforward that the representativeness of sampling, for both food and soils, covers a primary role in the robustness of the traceability models.With this aim, the issue of selecting a set of representative, informative and different soil samples is tackled. In this case, the goal is not obtaining a set of soil samples uniformly spanning the territory, since the planning of a punctual sampling of the whole district of Modena is not feasible considering the total number of affordable samples, but rather choosing a representative set of vineyards where the soil samples can be located. Thus, all the vineyard-registered producers of the district of Modena were considered and different variables (geological features of the soils, winegrowing coverage, grape varieties, yearly production of the farms, etc.) were handled with Experimental Design (DoE) techniques, to simultaneously take into account the different information in order to achieve a sustainable and rational site sampling. In particular, D-optimal onion design was chosen since it is widely used for mapping and planning purposes, hence it permits to achieve the maximum coverage and uniformity of the selected samples in the whole domain. An efficient mapping of the geographical region has been obtained ensuring coverage of farms characterized by the main grape production and insisting on soils with different geological features.